Clarify DSCP #174
Clarify DSCP #174
Conversation
draft-ietf-masque-connect-ip.md
Outdated
| When a connection is congestion-controlled, marking packets with different DSCP | ||
| can lead to reordering between them, and that can in turn lead the congestion | ||
| controller to perform poorly. If tunneled packets are subject to congestion | ||
| control by the outer connection, they need to avoid carrying different DSCP |
There was a problem hiding this comment.
Do we want to clarify that the possible solutions to this are multiple HTTP/3 connections with independent CONNECT-IP sessions?
There was a problem hiding this comment.
Isn't that already there further down in that paragraph?
There was a problem hiding this comment.
Yes, you're right, but I had to read it 3 times to notice. I think it gets missed with the note to scope on particular DSCP values. Maybe let's add a newline and separate that out into its own paragraph, since it's kind of a separate thought?
draft-ietf-masque-connect-ip.md
Outdated
| markings to not disrupt the congestion controller. In this scenario, the IP | ||
| When a connection is congestion-controlled, marking packets with different DSCP | ||
| can lead to reordering between them, and that can in turn lead the congestion | ||
| controller to perform poorly. If tunneled packets are subject to congestion |
There was a problem hiding this comment.
I would suggest to clarify that it is the "... in turn lead the HTTP connection's congestion controller to perform poorly."
draft-ietf-masque-connect-ip.md
Outdated
| can lead to reordering between them, and that can in turn lead the congestion | ||
| controller to perform poorly. If tunneled packets are subject to congestion | ||
| control by the outer connection, they need to avoid carrying different DSCP | ||
| markings to prevent this situation. In this scenario, the IP |
There was a problem hiding this comment.
"carrying different DSCP markings" is over simplifying it. It is carrying DSCP that are not equivalent in forwarding behaviour that is the issue.
| the tunneled packets need to be treated equally regardless of their DSCP | ||
| markings to not disrupt the congestion controller. In this scenario, the IP | ||
| When an HTTP connection is congestion-controlled, marking packets with | ||
| different DSCP can lead to reordering between them, and that can in turn lead |
There was a problem hiding this comment.
It not only reordering. Also getting different delay samples can break delay based cc.
There was a problem hiding this comment.
Do we need to go that much into detail?
There was a problem hiding this comment.
I think you simplification works fine here. As the relevant aspect cut out appear to covered in later sentences.
There was a problem hiding this comment.
I know this has been merged now but this comment of mine was addressed. reordering alone is not the problem.
Fixes #173